I'm thinking about adding crossover traps at the ends of the mains, on a 2-pipe vacuum system...

ted_p

...... to help get the steam to the farthest parts of the building quicker (similar to what Gerry Gill and Steve Pajek describe doing with added main vents on conventional systems in Dan Holohan's book "The Lost Art of Steam Heating Revisited"). I found this website and Dan’s books just a few months ago, and have learned more about steam systems in those months than I had in the 16+ years prior. It currently takes more than 4 minutes from the time the takeoff gets hot, for the steam to make it to the ends of both mains. I know if I can shorten that transit time, I can heat the building just as well with shorter burns; and the Heat-timer will shorten the burns automatically, because the system-sensor will be satisfied sooner.

I'm soliciting advice, suggestions and comments on this idea. This would be done as part of my off season projects, which already include replacing crazy-oversized F&T traps, and installing orifices in the radiator inlets.

A bit of background:
This is in an eight story, 72 unit apartment building in Northwest DC built in the early 40's. It has 2 steam mains of un-equal length, that run in opposite directions around the perimeter of the building in the sub-basement and basement levels (the sub-basement is only partial) feeding 282 cast iron radiators. With the exception of those in the bathrooms, all radiators are located against exterior walls, underneath windows. Both mains end several feet apart, under the 08 tier of apartments; the 08 tier living rooms each have two radiators, one fed off of each main. None of the risers are dripped.

It was originally a Dunham Vari-Vac system, which was converted to an ordinary vacuum return system, with a Shipco LRV (condensate/vacuum pump and receiver) and a Heat-Timer EPU (with a very poorly placed system sensor) in the late 80’s. In the early 90’s, the giant round steel boiler was replaced with a Weil-McLain 1288 series 1.

I took over managing the building in 2003, and have been learning about steam as I go. Rebuilt all the radiator traps (Dunham Bush, and Mepco 1E’s) the first time in ’05. In ’08 I replaced the Heat-Timer EPU with a Heat-Timer SRC Platinum, and relocated the system sensor to the top of the last supply riser with a crossover trap (to prevent the tenant in 808 from ruining me by closing the radiator valve). Installed Danfoss TRV’s on all the larger radiators in ’18 and on as many of the smaller (kitchen and bath) radiators as we could last year (had to skip some that are in such tight places that there simply isn’t room for the valve). I’ve been using temperature loggers since before I changed the Heat-Timer, to monitor air temperature, and on pipes and radiators to help with balancing and timing. This season the temperature difference between the warmest and coolest units is about 3 degrees, down from 10+ degrees when I started keeping track.

EBEBRATT-Ed

@ted_p

Sounds like your making a lot of progress.

As far as getting the steam to the ends of the Mains faster you can certainly use cross over traps if needed. What is venting the mains now? Anything? Or were they venting through the radiators?

depending on the length of the mains 4 min. may be good or bad

Pumpguy

Tell us more about the vacuum pumps and their performance.

What are the on and off settings for the vacuum switches?

How much time do the pumps take to cycle between the on and off vacuum switch settings?

What is the Shipco LRV unit's model number?

What is the temperature of the returning condensate?

Jamie Hall

As @EBEBRATT-Ed said, you're doing pretty well. Crossover traps will certainly help getting steam to the ends of the mains -- probably significantly -- since you will no longer be counting on assorted radiator traps to do the job.

Steamhead

Are the coldest units toward the ends of the mains?

ted_p

@EBEBRATT-Ed : There's a 3/4" Barnes & Jones H pattern F&T (vents 1.500 CFM at 3oz *) trap at the end of the short main, and a 2" Hoffman F&T (vents .950 CFM at 3oz *) at the end of the long main. That one is definitely being replaced this summer.

I've never measured the length of the mains, and it would be difficult; they start in the sub-basement, the jump up to the basement level, passing thru more than a few walls on the way. That said, the footprint of the building is not large (around 80 ft in the longest dimension), so they can't be terribly long. I'll try to add some kind of a sketch later today.

@Pumpguy : On at 3" off at 8"

Since we installed all the TRV's it's performing better than I'd ever seen; when the boiler's not firing, it cuts in at 3" and runs up to 8"on one pump in about six minutes, then stays off for maybe 20 minutes. Even when the boiler's firing, it can now hold well over than 3" of vacuum in the receiver. Had me baffled, until on my 2nd reading of Dan's book I realized what the problem was before the TRV's: Over the years a couple hundred of the original packless radiator valves had been replaced with cheap packed-valves that leak vacuum around the stem. When we replaced those valves with TRV's that leakage went away. And all those years I'd just assumed that there were lots of leaks in the piping.

Model 10LRV2-20-35

When it's cold, the return temp briefly goes above 170 during the morning boost, but stays below 150 the rest of the day. Here's a temperature graph from a few of the colder days in February (we had a really mild winter this year)



@Jamie Hall Thanks for the encouragement.

@Steamhead : The coldest units are on the top floor, but the steam goes up the risers pretty quickly. A week ago I had loggers on top of the same radiator in 303 and 803 (both were vacant at the time); the steam was getting to 303 less than a minute ahead of 803. I really think the difference is the heat loss thru the roof. It's apparent to me that the radiators were sized primarily to the windows, but when the old double hung wood windows were replaced in the 80's that screwed that up. One of the first things I realized when I started trying to balance the heat 17 years ago.

* The vent capacities are from @gerry gill and Steve Pajek's table in "The Lost Art of Steam Heating Revisited"

Noel

I once cared for 5 storys of a (previously bypassed) Vari-Vac system that had TRVs added to it. It was a classroom building and the mains fed 5 floors with vertical risers serving radiators with no horizontal piping on any of the risers. When the TRVs were added, there were no crossover traps added at the tops of the risers and in the shoulder seasons, the rooms would satisfy the TRVs before the steam would shut down on WWSD most days and the risers would cool down and condense. They would fill with water that would just hang there, pushed up by steam pressure and not flowing to the returns even when a room cooled and a TRV opened. Shutting off the steam would let the water fall and the cycle would begin again the next day.
The symptoms would be a no-heat call in a different stack of rooms each day. It took a while to identify what was going on. A crossover trap installed under the top radiator on each riser cured it.

dopey27177

You had a Dunham system. It is now a hybrid.
Do you still have an original Dunham zone valve.
Hope you do, because Dunham made the trim inside the correct size for the amount of steam the building needs. (Lets say your steam piping is 8" and the amount of steam needed to supply the building could come from a 6" valve, Dunham understood that reducing the valve size to 6" is costly because 2 8x6" reducers would be needed. So Dunham made the internal trim inside the valve 6")

Additionally, the zone valve had a micro switch on it. The micro switch was activated when the zone valve began to open, it started the vacuum pumps, and vacuum would be established immediately. Although there was no vacuum in the system both pumps came on and started the differential. When vacuum reached about 8" one pump would shut down.

If the system was tight a building could see as much as 26" vacuum at a low zone valve opening.

Remember, Dunham vari vac systems the zone vale floated in response to the temperature controller in the basement.

Unfortunatlly the best steam heating system ever invented was ruined.

You do not need crossover traps in the system. Have the vacuum pumps set to start up when the zone valve begins to open.

As far as installing orifices in the radiators they are already in the Dunham valves. see enclosure.

Remove the the TRVs and replace them with original Dunham valve.

If you have money in the budget have MEPCO come down and give you a price to restore the system to original.

The heating balance in the building is a combination of 2 things,

1. proper orifice setting in the radiator valves.
2. a panel that is designed to properly operate all the Dunham
mechanical equipment and work with signal from the outdoor
sensors, differential controller and indoor temp controller.

Here is a statement that I made to many of my customers when I was called to fix an old steam system, you had a Rols Royce and installed Voltswaggen parts.

Jake

ted_p

@Noel : Many thanks for sharing your experience; if a problem like that crops up, I'll know what to look for. Fortunately, we've had the full complement of TRV's in for an entire heating season now, and have had no such problems yet. Keeping my fingers crossed....

Regarding the crossover traps at the top of the risers, I already did exactly that 12+years ago on the riser where I put system sensor for the Heat-Timer, to ensure that the steam always makes it to the top, even when the tenant in the top floor closes the radiator valve.



This crossover trap is in a soffit, in the 7th floor unit, just below the 8th (top) floor radiator. You can't see everything in the pic because the hole in the soffit is pretty small. The system sensor is strapped to the riser, under the bulge of insulation at the bottom of the pic (mostly out of sight).

At the time I installed that, I'd never even heard of a "crossover trap". I just knew that I had to come up with some means to ensure that the steam would make it up to the sensor every time, even if the tenant in 808 closed the radiator valve, and that was the best solution that I could dream-up.

@dopey27177 :

Thanks for your suggestions.

The Dunham vacuum pump, differential controller, and the vast majority of the original packless radiator valves were long gone before I ever got involved. All that remained of the Vari-Vac stuff was the motorized valve and this old control panel (which was the only thing that made me aware that there'd ever been a Vari-Vac system there in the first place).



For years I'd assumed that that old panel was just Dunham's take on an ordinary outdoor-reset, cycling steam controller. Boy, was I wrong!

I only learned very recently what a Vari-Vac system actually is, from some old literature I downloaded from the Museum on this site, and am now quite intrigued by the concept; but not enough to pay what it would cost to put it all back. Particularly when I've got the cobbled together system we have now working pretty darn well. I just think that I can make it work better, and save some fuel, if I can get the steam to the end of the mains quicker.

Regarding the orifices I'm installing, these are primarily a means of doing away with radiator trap maintenance, which is a significant ongoing expense. In '05, when we went through all the radiator traps in the building the first time, I was buying Mepco discs for less than $14 a piece. Now, the best price I can find is over $25; that times 282 traps is money that I'd much rather spend on other things.

dopey27177

The crossover traps you are referring to are riser vents (High capacity vent valves).

After some reflection I believe the building was converted to a Dunham Varivac System because these riser vents are not needed in a vacuum heating system.

The vacuum pumps pull all the air out of the system long before
the steam would push the air to the steam traps.

Think of the pressure differential with out a vacuum pump, steam pressure 2 PSIG. With a vacuum pump 4" HG differential.
The 4" HG differential is equal to 4 PSIG.

With the vacuum pumps in place Dunham would not want the riser traps in place because a trap failure will send steam thru the return riser at a far greater rate and temperature than the steam leaving a radiator.

Each radiator in a Dunham system had and orifice that regulated the amount of steam to what the radiator had to supply in heat to the room, there fore the condensate temp may be hot and very little steam would leave the radiator thru a failed radiator trap.

So it seems to me the building has had its third conversion. Hope all works well in the future, its good to think outside the box.

Jake

Joe_Dunham

FYI Vari-Vac is still specified in new construction in NYC and the 5 boros. The latest panel is IoT. ALL the equipment is still available. TIP; vacuum leaks go undetected because the atmosphere is leaking INTO the pipe. However, vacuum leaks (lack of tightness) cause distribution problems. Picture drinking a soda through a cracked straw. You get more air than soda. A bad expansion joint, a broken run-out, numerous leaking packing, etc., will affect steam flow the same way. Close the main return to the vac Pump, run it on hand and see what you get. Should be 20" right away. Then open it and isolate the boiler. Run the Pump on hand and watch the vacuum gauge. Let it run until until it stops rising. That's the cold system vacuum. 20" or more is perfect, 15 is acceptable. Steam under 20" Hg is 150F or so. We run "tightness" tests with peppermint oil to locate leaks. You don't necessarily need to change all the TRV's to get a big benefit from Vari-Vac. and since the valve is there, it may be alot cheaper than you think. Continuous flow as opposed to on/off

Steam34

Joe_Dunham

ted_p

dopey27177 said:

The crossover traps you are referring to are riser vents (High capacity vent valves).

After some reflection I believe the building was converted to a Dunham Varivac System because these riser vents are not needed in a vacuum heating system....

@dopey27177 : I fear that I've not described things well, or the picture that I posted is misleading. There are no riser vents, or crossover traps on the risers, except for the single crossover trap (in the picture above) which I installed 12 years ago (in conjunction with the system sensor for the Heat-Timer). To the best of my knowledge, the only vent in the entire system is the one in the condensate receiver.

I'm quite sure that that it was originally a Vari-Vac, or at the very least some other kind Dunham system, because all of the original radiator traps are Dunham 1E's. It's easy tell which traps are original (and we still have a a bunch of them), because they connect to the radiator bushings with a left/right threaded nipples, instead of a union.

@Steam34 : I Like the pic. Is that in your system? I think I'm jealous.

@Joe_Dunham : Many thanks. I'll be back at the building the week after next, and will try the tests you suggested and report back.

Some things I can tell you about it's performance though: Early this season the condensate receiver started flooding because the water wasn't returning to the boiler quickly enough (it turned out the the diaphragms in both of the discharge valves were perforated), so when a big bolt of condensate returned both pumps continued running on float, long after the vacuum switches were satisfied. When this happened and the boiler wasn't firing, the vacuum would go above 15", and that was without isolating the boiler. I was astounded, as this was the same unit that struggled so long and hard just to reach 8", before we replaced so many packed radiator valves with TRV's. The tech from Shipco who helped me troubleshoot the condensate return issue also suggested that we replace the venturi plates because (based on the age of the unit) they're likely eroded to the point where it has significantly reduced vacuum production, so that's now on my list of off-season projects too.

But I've got a question about using this in a restored Vari-Vac system: This is a venturi-jet type vacuum pump, which I don't think is designed to produce the high vacuum that Dunham vacuum pumps do. Based on my recent experience, I know it'll pull 15" and most likely 20" of vacuum with the boiler isolated. But is doing that on a regular basis going to be like using a single-stage air-compressor to pump air at 160psi? A single-stage compressor in good shape can make 160psi air, but it runs a long time to produce darn little volume compared to a two-stage compressor half the size.

Joe_Dunham

100% this was Vari-vac from original construction. I know by the panel. Whats the address? I may even have the original orifice adjusting schedule. the manufacturer gave them all to us because we in NY/NJ have kept vari-vac alive.( When yo say diaphragm, sounds like the pump has hydraulically actuated discharge valves. they can be a nuisance). A jet pump design has been used by dunham and most other pump manufactures for many years. it is simple and reliable. it is adequate for you're application. Ventrite (formerly dunham) makes 3 different style pumps;
1- the very large pumps (over 65000 edr) use liquid ring pumps. not a jet pump design. not that common
2- A model 'E" pump (jet style) . basically a tank on top of a tank. The upper tank has the jet pumps and the lower accumulator tank has the condensate pumps. this is not a great design. it requires make up water at the top tank. make up water rots the tank. they also stick an aquastat to "cool" the top tank. that makes it rot faster, wastes water and is a band aid for passing traps. the lower pumps are subject to vacuum. no good. leaking check valve are a loss of vacuum. this is usually speced when an engineer want a cast iron accumulator (which is ironic because they don't realize the make up water will rot the TOP tank out.
3- the third design is the C5 (jet style). it has an integral accumulator tank. (thats the one in the picture) all four pumps are in the hurling tank. getting rid of condensate is two steps. the air pumps suck air and water out of the accumulator tank. the air is blown out the vent and the water, as the level increases, is pumped out by the condensate pumps. there is NO make up water. so the two condensate pumps take the place of your diaphragm valves. This is the the best design. simple, reliable, easy to troubleshoot and maintain. it can also be bought as a combo boiler feed vacuum pump. very handy.

alot of times a cold system vacuum will be lower than the vacuum you see when the building first begins to heat. the steam closes the thermostatic elements and the system supply side is somewhat isolated from the return, making the system "smaller" and isolating supply side leaks. but with true vari vac, most weather conditions you will have vacuum on the SUPPLY as WELL as the return. so tightness is important all the way across the system. the vacuum is varied in order to vary the TEMPERATURE of the steam. like hot water reset

Joe_Dunham

this was you original equipment pump. model "C" up to 30000EDR

Joe_Dunham

the C5